jt burke
Marion, OH
C&O and Chessie System Modeler
www.marionmodelrailroadclub.org
You said the magic words. Sorry, no duck, and no hundred dollars, but "Capacitive Discharge" is the way to go. I use an old, old (I mean, like, old. Groucho was still on TV when I got this thing) train transformer. I use the AC terminals, about 12-14 volts, and run them into a home-built CD circuit.
Since you're running 4 machines at once, I'd recommend beefing up the big capacitor. That means more microfarads. It will hold more of a charge, and will be able to kick out more power all at once.
The input supply will then determine how fast you can re-charge your CD circuit. With mine, I need to wait a couple of seconds between throws. If you have a beefier supply (more amps) then your circuit will recharge faster.
It takes an iron man to play with a toy iron horse.
Musepro wrote:Hello all!I am building a layout that uses Atlas under-the-table switch machines and need to build a DC power supply to power them. I'm using diode matrix and capacitor discharge circuits. I have two books on model railroad wiring and have read differing view points in these and on the web. I've heard anywhere from 6VDC to 20VDC and from 2AMPs to 6 AMPs is needed. I will be powering no more than 4 machines at a time (used in staging). The Radio Shack 18V, 2AMP tansformer stated in one book is out of production. Radio Shack does have a 12.6V, 3AMP transformer and a 25.2V, 2AMP model in stock. I thought I could parallel the secondaries on a pair of the 12.6V, 3AMP transformers if I had to.Has anyone built a DC power supply for their twin solinoid machines lately? What are your thoughts of the required voltage and amperage I will need for 4 simultaneous machines using capacitor discharge? Thanks for your help!jt burkeMarion, OH
welcome aboard, JT.
You don't need to double up - a single 12.6v/3amp filament transformer (same one I use) will charge the capacitor(s) of a CD circuit to within a close handshake of 18vdc, which is probably about as much voltage as you want to shoot to those wimpy Atlas machines. (Much more and the plastic parts will be at serious risk of self-destruction.)
The key to throwing several turnout actuators at a time is total capacitance - which is also the key to recharging speed. I've found that it takes about 1500 microfarads per machine to kick over my old KTM rocksmashers, and about 1000 microfarads to throw Rix machines. One of each will give positive results on 2500 (crossover at the entrance to a hidden staging yard.)
Recharging time is only an issue if you can throw switches as fast as a speed-typist hits keys. Using a hot probe (look, Ma, no switches) I've found the recharge time to be less than the time it takes me to move the probe to the next stud.
Chuck (modeling Central Japan in September, 1964)
Musepro wrote:Chuck,So a single 12V/3A transformer will do the job? Have you any recommendations on capacitor size for the Atlas machines? I'm not familiar with the KTM machines but I assume the Atlas machines wont need anymore than the Rix. I was planning on installing a capacitive discharge for each machine. Thanks for your input!jt burkeMarion, OH
No, a single CD circuit should do the job for your whole layout.
MisterBeasley wrote: Musepro wrote:Chuck,So a single 12V/3A transformer will do the job? Have you any recommendations on capacitor size for the Atlas machines? I'm not familiar with the KTM machines but I assume the Atlas machines wont need anymore than the Rix. I was planning on installing a capacitive discharge for each machine. Thanks for your input!jt burkeMarion, OHNo, a single CD circuit should do the job for your whole layout.
The switch machines are connected to the power supply by a momentary switch; there is no such thing as "stuck on". The power throws a solenoid to one end of travel with one connection and to the other end of travel with the other connection (there is a third, common power connection). If you leave power on either of the two connections; you burn up the switch machine. You need a CD system powerful enough to throw four switches at the same time. Forget about using individual caps (which would not even be connected to anything until you used the momentary switch to throw turnouts).
.
Alan_B wrote:The switch machines are connected to the power supply by a momentary switch; there is no such thing as "stuck on". The power throws a solenoid to one end of travel with one connection and to the other end of travel with the other connection (there is a third, common power connection). If you leave power on either of the two connections; you burn up the switch machine. You need a CD system powerful enough to throw four switches at the same time. Forget about using individual caps (which would not even be connected to anything until you used the momentary switch to throw turnouts)..
What could 'stick' is a failed momentary contact pushbutton or switch - and the ones most likely to fail are the Atlas switches and those really cheap Radio Shack microbuttons. If you use either, the question isn't whether they will fail, it's when.
I avoid the problem by using two momentary contact systems that are just about failureproof, one dirt cheap and the other rather expensive:
As to your capacitor question, I would experiment with a 2500 microfarad capacitor, then add capacitance in 1000 mufd increments if necessary to secure reliable operation.
tomikawaTT wrote:Shorting rotary switch - I use a 4p3t model, with one pole to throw the switch and the other two available for interlocking and interpanel communication. Looking ONLY at the throw system, there is NO connection to the pole itself. The four throw connections are, in sequence, turnout normal route coil, turnout hot, turnout reverse route coil, off. Turning the rotary switch knob 60 degrees causes the pole to momentarily short the hot connection to the coil connection adjacent, but the switch detent won't permit the pole to remain in the shorted position. Once again, the switch goes in the panel track diagram, with the knob position indicating which way the points were most recently thrown.
Shorting rotary switch - I use a 4p3t model, with one pole to throw the switch and the other two available for interlocking and interpanel communication. Looking ONLY at the throw system, there is NO connection to the pole itself. The four throw connections are, in sequence, turnout normal route coil, turnout hot, turnout reverse route coil, off. Turning the rotary switch knob 60 degrees causes the pole to momentarily short the hot connection to the coil connection adjacent, but the switch detent won't permit the pole to remain in the shorted position. Once again, the switch goes in the panel track diagram, with the knob position indicating which way the points were most recently thrown.
Hi, I also use a version of the stud-&-probe method. For 'studs', I use 2-56 Stainless steel bolts; Stainless steel does not weld itself to the meter probe I use to select which T/O motor to throw. The bolts are cheaper than p-buttons, too!
For power supply, I think it would be much simpler for you to use the AC output of one of your power packs to feed your CD unit. At rest, the units don't draw much power. Mine is a simple homebrew junkbox circuit that gives a good whack to the T/O coil initially and then fades to about 2 volts as the capacitor dies. Now I don't fry the coils. Atlas machines don't seem to draw the power the Rix ones do, but they overheated very quickly if I didn't get off the push-buttons I used to use, hence the meter-probe method.
This works well for me, give it a try yourself. Have fun, George
Musepro wrote: Alan_B wrote: The switch machines are connected to the power supply by a momentary switch; there is no such thing as "stuck on". The power throws a solenoid to one end of travel with one connection and to the other end of travel with the other connection (there is a third, common power connection). If you leave power on either of the two connections; you burn up the switch machine. You need a CD system powerful enough to throw four switches at the same time. Forget about using individual caps (which would not even be connected to anything until you used the momentary switch to throw turnouts).. I understand the wiring. I actually meant the normally open push buttons sticking closed or failure due to excess current. This does happen, but a cap discharge unit will prevent machine burn out because the cap cant be recharged if the circuit is left closed (supposing the DC supply is small enough).Just as I've read elswhere, there are a lot of ideas here too. I guess they all work, was just wondering if any of them were more proven than others and about how much current/voltage I'd need for 4 Atlas machines. That was my orginal question and really my only question so far; how much power do I need for 4 Atlas machines with a discharge circuit and about what size of caps?jt burkeMarion, OH
Alan_B wrote: The switch machines are connected to the power supply by a momentary switch; there is no such thing as "stuck on". The power throws a solenoid to one end of travel with one connection and to the other end of travel with the other connection (there is a third, common power connection). If you leave power on either of the two connections; you burn up the switch machine. You need a CD system powerful enough to throw four switches at the same time. Forget about using individual caps (which would not even be connected to anything until you used the momentary switch to throw turnouts)..
As Alan pointed out, a C-D supply for each turnout is just not practical. Don't go there. Yes, pushbuttons can weld together under high current--if you use the cheap ones. As someone else said, stay away from the R.S. cheapie micro switches. Use switches rated at 2 or 3 amps. The chances of one of them welding together is extremely small. The power to throw the turnouts comes from the capacitor, not from the transformer. The transformer is there to re-charge the capacitor between uses. A 2 amp transformer is plenty. As someone suggested, start with around 2,000 mfd and if that doesn't provide enough of a jolt, add another 1,000 mfd and even another 1,000 until your 4 turnouts snap nicely. Two words of warning. 1) The jolt from 4 or 5,000 mfd may be more than one single turnout coil can handle. 2) Make very sure that you connect positive to positive and negative to negative when joining capacitors together. Otherwise you could have a very nasty, messy and painful explosion.
..... Bob
Beam me up, Scotty, there's no intelligent life down here. (Captain Kirk)
I reject your reality and substitute my own. (Adam Savage)
Resistance is not futile--it is voltage divided by current.
I had not thought of point 1 above (edited to enlarged and bold). I built a CD unit to throw two Atlas snap switches at a time (several crossovers in my layout); it does work fine with just one switch. I used parts-on-hand and just added caps until I got reliable operation. I think that I ended up with about 3000 mfd; however the CD system is now mounted under my bench and I can not read the cap values for verification.
Do you really need to throw 4 switches at one time? Complicated, diode matrix, switching systems do not work well with snap switches; but work great with stall motor machines like the tortoise.
The problem with "one machine stuck on" which inhibits recharging of the capacitors shouldn't be a concern. That's like NASA fail-safe, man-rated system engineering. If one switch machine is "stuck on," (because of the pushbutton, not the switch machine, of course) then you need to stop running and fix the problem. Without a CD circuit, your nose is likely to notice it first, as the coils fry themselves into oblivion. The pass-through current limitation (the input resistor) of the CD circuit actually protects the switch machines in this case, and you may get away without frying it.
I would advise using a common CD circuit for many switch machines. First of all, it's a lot cheaper and easier, and second, it will give you a way to detect a problem if one button sticks. If you use individual circuits, then you won't notice anything is wrong until you go back to the stuck turnout, but with a single circuit for all, you'll see you've got a problem the next time you try to throw any turnout. I'd say the ability to detect the error is more important than the ability to keep operating after it's happened.
Alan_B wrote: As Alan pointed out, a C-D supply for each turnout is just not practical. Don't go there. Yes, pushbuttons can weld together under high current--if you use the cheap ones. As someone else said, stay away from the R.S. cheapie micro switches. Use switches rated at 2 or 3 amps. The chances of one of them welding together is extremely small. The power to throw the turnouts comes from the capacitor, not from the transformer. The transformer is there to re-charge the capacitor between uses. A 2 amp transformer is plenty. As someone suggested, start with around 2,000 mfd and if that doesn't provide enough of a jolt, add another 1,000 mfd and even another 1,000 until your 4 turnouts snap nicely. Two words of warning. 1) The jolt from 4 or 5,000 mfd may be more than one single turnout coil can handle. 2) Make very sure that you connect positive to positive and negative to negative when joining capacitors together. Otherwise you could have a very nasty, messy and painful explosion.I had not thought of point 1 above (edited to enlarged and bold). I built a CD unit to throw two Atlas snap switches at a time (several crossovers in my layout); it does work fine with just one switch. I used parts-on-hand and just added caps until I got reliable operation. I think that I ended up with about 3000 mfd; however the CD system is now mounted under my bench and I can not read the cap values for verification.Do you really need to throw 4 switches at one time? Complicated, diode matrix, switching systems do not work well with snap switches; but work great with stall motor machines like the tortoise.
MisterBeasley wrote:The problem with "one machine stuck on" which inhibits recharging of the capacitors shouldn't be a concern. That's like NASA fail-safe, man-rated system engineering. If one switch machine is "stuck on," (because of the pushbutton, not the switch machine, of course) then you need to stop running and fix the problem. Without a CD circuit, your nose is likely to notice it first, as the coils fry themselves into oblivion. The pass-through current limitation (the input resistor) of the CD circuit actually protects the switch machines in this case, and you may get away without frying it.I would advise using a common CD circuit for many switch machines. First of all, it's a lot cheaper and easier, and second, it will give you a way to detect a problem if one button sticks. If you use individual circuits, then you won't notice anything is wrong until you go back to the stuck turnout, but with a single circuit for all, you'll see you've got a problem the next time you try to throw any turnout. I'd say the ability to detect the error is more important than the ability to keep operating after it's happened.
I use 12.6 volts at 2 amps and 4700uF to throw pairs of atlas machines in hidden staging with no problems (I prefer the tortoise for visable turnouts). You should be fine with the above suggestions. Recharge time is about three seconds due to the size of the series resistor. I also use center off five amp mini toggle switches.
Good luck,
Karl
The mind is like a parachute. It works better when it's open. www.stremy.net